Can't name my columns as dates - r

Question

I have a range of dates where some products were bought. I create a sort of a pivot table relating the products and the dates, but there are dates where nothing was sold. I can find the missing dates and even add them to the main data frame, the problem is that instead of keeping the date format, they adopt the integer format (with the integer being the distance to origin) and I can't order them. The code I'm using is this:
upper.bound<- paste("01", month[1], 2013, sep="-")
lower.bound <- paste("30", month[4], 2013, sep="-")
dates <- seq(as.Date(upper.bound, "%d-%m-%Y"), as.Date(lower.bound, "%d-%m-%Y"), "days")
diff <- setdiff(dates, as.Date(colnames(export_f_ub), "%Y-%m-%d"))
len <- dim(as.matrix(diff))[1]*11
aux <- data.frame()
aux <- seq(0,0,length.out=len)
dim(aux) <- c(11, dim(as.matrix(diff))[1])
col_dates <- as.Date(diff, origin="1970-01-01")
colnames(aux)<- c(col_dates)
This was a tryout to set the matrix to zeros and then bind it to the main one. But this doesn't work, as in the result I get the column names as numeric. Here's a screenshot of the console:
Console log

Answer 1

I've never seen someone try to assign a Date vector as column names of a matrix. Dimension names must always be character strings, so in general this is not something you should be doing.
That being said, in terms of effect, the intuitive expectation would be that the column-name-assignment machinery in R would at some point coerce the Date vector to character along the lines of as.character(), and thus you'd get the text representation of the dates, rather than a stringification of their underlying double values.
Calling `colnames<-`() on a matrix eventually calls `dimnames<-`() which drops into the C code by running .Primitive("dimnames<-"). I haven't really looked into the C implementation, but we can guess that at some point it pulls out the double values underlying the Date vector, coerces them to character, and that's why you end up with numbers as your column names.
The correct approach here is to call as.character() yourself when assigning the names:
col_dates <- as.Date(c('2013-06-03','2013-06-04','2013-06-05','2013-06-06','2013-06-08','2013-06-22','2013-07-07','2013-07-08','2013-07-11','2013-07-13','2013-07-23','2013-07-25','2013-07-26','2013-08-27','2013-09-03','2013-09-04','2013-09-05','2013-09-06','2013-09-07','2013-09-09','2013-09-10','2013-09-11','2013-09-13','2013-09-14','2013-09-15','2013-09-16','2013-09-18','2013-09-20','2013-09-21','2013-09-22','2013-09-24','2013-09-30'));
aux <- matrix(0,11L,length(col_dates));
colnames(aux) <- as.character(col_dates);
aux;
## 2013-06-03 2013-06-04 2013-06-05 2013-06-06 2013-06-08 2013-06-22 2013-07-07 2013-07-08 2013-07-11 2013-07-13 2013-07-23 2013-07-25 2013-07-26 2013-08-27 2013-09-03 2013-09-04 2013-09-05 2013-09-06 2013-09-07 2013-09-09 2013-09-10 2013-09-11 2013-09-13 2013-09-14 2013-09-15 2013-09-16 2013-09-18 2013-09-20 2013-09-21 2013-09-22 2013-09-24 2013-09-30
## [1,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [2,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [3,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [4,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [5,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [6,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [7,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [8,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [9,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [10,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [11,] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
But you should be aware that the column names do not retain the Date class or internal representation (referring to the double values); they are pure character strings. If you want to recreate the Date vector from the column names, you'll have to run them through as.Date().
And by the way, dim(as.matrix(x))[1] for a vector x is an unnecessarily roundabout way of getting length(x).